"This book, which is a collection of articles by experts, serves a vital need for a good general book on the subject of observational studies in medicine. The randomized control trial is not the only method of analysis, but observational studies have their special problems. This book concentrates on method such as propensity scoring and instrumental variables, pointing out the advantages and disadvantages of each. The technical level varies from chapter to chapter, but is generally fairly accessible. The chapters all have extensive references, and each has copious SAS code, with comments and output, to illustrate key ideas. If you analyze observational data in the medical field, you will want this book."

Peter Flom
Independent Statistical Consultant
Peter Flom Consulting, LLC